A philosophical essay on artifacts and norms

(1)

A philosophical essay on artifacts and norms

Citation for published version (APA):

Vaesen, K. (2008). A philosophical essay on artifacts and norms. Technische Universiteit Eindhoven. https://doi.org/10.6100/IR636332

DOI:

10.6100/IR636332

Document status and date: Published: 01/01/2008 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

(2)

A Philosophical Essay

on Artifacts and Norms

Krist Vaesen

‘Wonder en is

Ar tifacts and N or ms K rist Vaesen Simon S tevin S er ies in the P hilosophy of Technolog

gheen wonder’

virtues governing scientific practice. Sound theories, for instance, should be falsifiable, empirically adequate, and/or simple. Analyzing these features helps to explain what science is or should be, and to describe how science has evolved or is still evolving.

A similar analysis is missing for technological practice. We can reasonably expect though that a study of the norms governing technology - e.g. product norms, technical standards, guidelines for design and use - will likewise prove useful to clarify what engineers and users do and why technical artifacts are the way they are. My thesis takes up this point. It provides a taxonomy of norms endemic to tech-nology and a philosophical analysis of the normativity involved. Arguments are supported by empirical data, such as case studies, evidence from cognitive psycho-logy, and research on design methodology. In this way my research addresses what actually goes on in technology and engineering.

(3)

A Philosophical Essay on Artifacts and Norms

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificus, prof.dr.ir. C.J. van Duijn, voor een

commissie aangewezen door het College voor Promoties in het openbaar te verdedigen op dinsdag 9 september 2008 om 16.00 uur

door

Krist Vaesen

(4)

prof.dr.ir. A.W.M. Meijers

Copromotoren: dr. M.J. de Vries en

(5)

(6)

Rector Magnificus, voorzitter

Prof. Dr. Ir. A.W.M. Meijers, Technische Universiteit Eindhoven, promotor Dr. M.J. de Vries, Technische Universiteit Eindhoven en Delft, co-promotor Dr. W.N. Houkes, Technische Universiteit Eindhoven, co-promotor

Prof. Dr. S.O. Hansson, Royal Institute of Technology, Stockholm, Sweden Prof. Dr. Ir. J.E. van Aken, Technische Universiteit Eindhoven

Dr. H.J. de Vries, Erasmus Universiteit Rotterdam Dr. M.P.M. Franssen, Technische Universiteit Delft

Research for this thesis was made possible by the Netherlands Organization for Scientific Research (NWO), project number 360-20-110.

c

Krist Vaesen, 2008.

Vaesen K.

A Philosophical Essay on Artifacts and Norms. Weidestraat 22, B-2600 Antwerpen

Email: k.vaesen@tue.nl ISBN: 978-90-386-1340-6 ISSN: 1574-941X

(7)

(8)

(9)

Never use epigraphs—they will only kill the mystery in the piece!

Gebruik geen motto’s, want die zijn dodelijk voor het mys-terie van het stuk!

(10)

(11)

Acknowledgements

So there you are, reading the first lines of Krist Vaesen’s doctoral thesis. Perhaps you are the kind of reader that lets his finger pass quickly over the lines, looking for the chain of strings he is trained to recognize, the letters making up his name. The most pressing question of the book: Am I in it?

You read the author write: yes, you are. You must be. He owes you you to surface.

Somewhere in between the Chapters You Will Never Read, and the Lines Above

That Made You Curious, you should find yourself. For instance, among the

good-natured Characters That Have Guided The Author’s Head And Hand1_{(these might}

in fact figure prominently in a sequel called 1001 Contributions To This Book). But don’t panic if you’re not one of them; may be you were one of the Author’s Sparring

Partners2, providing for jokes and trips and joint-cutting story lines? Or it could be

that you took up the role of a Person Making The Author’s Stays In His Room

Enjoy-able3_{, a category overlapping to a large extent with the category of Fellow-Authors}

Sharing The Thrills And Pains Of Autorship4_{? If you’re brown-haired, thin and into}

Fantasy, you might have been the Author’s Spell-Check; in case you’re increasingly non-haired, descreasingly thin, and German, you must be the Remarkable And

Re-markably Funny Mind That Would Look Like A Round Square If Uncovered5_{. If not}

all of these, you might be endowed with an Indian epitheton, like Crazy Squash6_,

Cosy Chat7_{or Pleasant Floor}8_.

1_{These people are: my promotor, Anthonie, and my supervisors Marc, Wybo and Jesse.} 2_{Melissa and Giacomo.}

3_{Jesse, Melissa, Giacomo, and Mechteld.} 4_{Melissa, Giacomo, Mechteld, and Auke.} 5_Christian

6_{My squash-partner Lambèr.} 7_{Rianne and Mieke.}

(16)

By the way, some of the first ideas for this book were put on paper in the Country

Of Country; there, a Hospitable Gentleman Of Dazzling Sharpness10put the author on the right track.

Oh no!, the oldest characters might think11_{, were am I, when do I enter the}

scene? — Here, replies the author appeasingly, here you are, in the Part Of Last But

Not Least: the Friends That Brighten Up The Author’s Life12_{, a Family Of Unflagging}

Support13_{, and, last but not least: The Author’s Love Of His Life}14_.

You see, you were in it! And you should be thanked for that! By the grace of your presence, the lines above could be populated, and the chapters to come could be filled. The authors owes you his gratitude, for if not for you, he would still be

The Author Of Empty Pages.

9_{The Netherlands Organization for Scientific Research (NWO).} 10_{Professor Jonathan Dancy.}

11_{Old qua being a character, that is.}

12_{It is close to impossible to give a list; but friends will certainly know that they are.} 13_{Especially the Author’s Mother and Father; but also the rest of his family and his in-laws.} 14_{The dear, dear Marjolijn.}

(17)

(18)

(19)

Chapter 1 Introduction

1.1 A Chinese experiment

As an appetizer, take a look at the figure below. Take notice of the fact that the strange looking device was photographed in the business high-rise area of Beijing. In accordance with the colors of the Chinese national flag, it is colored yellow (lighter shades of gray in the picture) and red (darker shades of gray). Assuming most of us have not encountered a similar object before, the obvious question is: What is it for?

To answer the question, we might look at the things the object affords. Going on appearances, for instance, the black spots in the picture are handles one is supposed to take hold of. Another clue is the height of the object, which amounts to 80 cm. Given this fact we can exclude some hypotheses; for one thing, it is too large to be of use on a sink top. For that matter, the thing was photographed in the streets, a fact that, on the assumption it was found in its normal context of use, also counts against the kitchen appliance hypothesis.

Although these considerations are useful, a more accurate interpretation would require us to zoom in on the little notice just below the alleged handles. Since the human eye will fail here, I have included a blow-up in figure 1.2. It appears the device is a piece of fitness equipment; in particular, it is supposed to strengthen liver and spleen, and to extend the Shao Yang channels of both sides of the chest. Moreover, the protuberances on the round plates are meant to massage your feet whilst working out. The instructions of how to use the device are straightforward:

(20)

Figure 1.1: A Chinese artifact

stand on a round plate, hold the handles tight with both hands, and turn the waist left and right (see sketch in figure 1.2).

Given this information, one can explain some other salient features of the ob-ject. Its height for instance: presumably the equipment is adjusted to the average height of the average (Chinese) passer-by. For purposes of strength and stabil-ity, furthermore, the designer has opted for a metal construction, rather than for one of cardboard. And perhaps the colors red and yellow serve a purpose as well: they might well correspond to the aesthetic preferences of the average (Chinese) pedestrian, or even bear witness to some political maneuver—patriotic uplifting, for example.

Additionally, the producer deemed it appropriate—or was forced by (contract) law—to provide for a warning notice. As a measure of safety, the sign recom-mends weak and ill people to consult a doctor before using the equipment, and says children are allowed to put it to use only under supervision of adults.

(21)

1.1. A Chinese experiment

(22)

there is a—perhaps latent—need for fitness equipment in the streets of the high-rise business area of Beijing. So what the artifact is for points to yet another fact, that is, to the reasons for its being there in the first place—or more confusingly, it broaches the question: why is there this object that is for this-and-that? Perhaps the artifact attests to current public health policy; perhaps government authorities acted upon the news of an alarming increase in spleen deficiencies amongst hard-working businessmen.

In sum then, one simple, but indeed somewhat weird artifact appears to raise a whole set of questions, questions one tends to lose sight of when confronted with more familiar objects, such as tables, spoons, cars, drain pipes, and whatnot. And many of these phenomena are in need of clarification. Indeed, we have theories about why biological items are the way they are, or closer to philosophy—in par-ticular the philosophy of science—about why scientific theories are the way they are; but surprisingly, such account is lacking for most of the things that make up our pervasively artifactual environment.

1.2 An instructive parallel

Consider the parallel to the philosophy of science more thoroughly. One of the main concerns of the philosophy of science is the establishment of criteria for scientific or rational knowledge, and a wide variety of such desiderata has been proposed. For the Babylonians, for instance, the “science” of astronomy was eval-uated for its predictive power, which means a good account was to give accurate predictions of the movements of celestial bodies. The pre-Ptolemaean Greeks, on the other hand, were more interested in causal patterns, that is, in explanatory

power. Much later, logical positivists claimed that scientific knowledge should be

assessed in terms of verifiability; through a comparison of the theories of Freud and Einstein, Popper considered falsifiability as the most salient criterion distin-guishing science from pseudo-science.

Additionally, an equally wide spectrum of criteria has been proposed to account for the soundness of a scientific theory (so “after” it has proven to be scientific, as opposed to pseudo-scientific). To give you an impression, I invite you to take a look at the table below. It presents a set of epistemic virtues1 scientific theories

1_{Usually, the term “epistemic virtues” refers to a cognizer’s intellectual virtues (cfr. Ernest Sosa’s}

(23)

quali-1.2. An instructive parallel

Boyle consistence, coherence with background knowledge, simplicity, natural/non-artificial, must lead to new testable predictions Peirce empirical support, explain as many

phenomena as possible, triangulation, simplicity, inexpensive to test, inertia, natural, initial credibility

Quine simplicity, generality, inertia, refutability Goodman coherence with research tradition, initial and

intuitive appeal, coherence with other “stories” about our world

Kuhn accuracy, simplicity, internal and external consistency, range of application, fruitfulness, comprehensive

van Fraassen explanatory power, range of application, practical usefulness, capacity to reconcile distinct disciplines and research traditions, precision, computability of a manifold of quantitative parameters, testability McMullin logical consistence, coherence, causal

specificity, coherence with other disciplines, innovative, comprehensive, cumulative, unification

Longino empirical adequacy, innovative, ontological heterogenous, complexity of the proposed relations, contributing to the distribution of power, usefulness regarding the fulfillment of human needs

Table 1.1: Epistemic virtues as proposed by different authors

should meet, as proposed by several authors (adapted from van Brakel, 1998). Apparently, these virtues are manifold, and indeed, there is no consensus at all which of them are essential as opposed to peripheral. This is not to say, how-ever, that the efforts of people like Boyle, over Quine and Kuhn have been futile. Perhaps, in a normative sense, their proposals have not had much impact on sci-entific practice itself; but at the least, these virtues give us a tool to explain what science is according to current or past practices, or to describe how science, post factum, has evolved or is still evolving. To give only one example, Putnam (1992)

(24)

develops a convincing argument showing why Einstein’s theory of gravitation was accepted and Whitehead’s alternative rejected, fifty years before anyone thought of an experiment to decide between the two. The author can do so precisely by invok-ing two epistemic virtues, namely simplicity (Einstein’s theory seemed a “simpler” way to move from Special Relativity to an account of gravitation than Whitehead’s) and conservatism (momentum was redefined by Einstein so that the Law of Con-servation of Momentum could be conserved in elastic collisions).

Moreover, these epistemic virtues may direct us to what we expect scientists to do, just like in ethics the Right (i.e. what ought to be done) can be understood in terms of the Good (i.e. what is valued). If we think our scientific theories ought to be empirically adequate, this would require scientists to follow certain procedures, say, working with a sufficient number of observations; if we would consider practical usefulness as primordial instead, we presumably hope scientists not to do science from the armchair.

Important to note is that we can let these notions do both normative and de-scriptive work. In the former case, we endorse or question the values at stake ourselves, and formulate recommendations from there on. Or we can take a more detached stand; irrespective of our own commitments, we make an inventory of the epistemic virtues prevailing at time t, and study their impact on scientific prac-tice. In this case, we make no evaluative judgments, we merely pursue descriptive ends.

Let me at last clarify the parallel to the philosophy of technology. To gain in-sight in how science proceeds, I argued summarily, it rewards to study the stan-dards the end-products of science (viz. scientific theories) are set against. Now, intuitively—and I will spend more effort to put this intuition on more stable ground—there seems no reason to expect that a similar method would not work for technology. If so, we would approach technological practice by studying the standards its end-products are supposed to meet. I guess it makes little sense to claim that artifacts should be empirically adequate, comprehensive or consistent; but what then is it we expect them to conform to?

With this the key issue of my dissertation is on the table. The overall question my research is aimed to answer therefore reads: what kinds of norms govern

techno-logical practice? In line with the above, I will take artifacts as a starting point. To

explain why artifacts are the way they are, I will attend to the demands we put on them; so, returning to our Chinese experiment, we expect the shape and form and

(25)

1.3. Expected outputs

responses of the fitness device to be clarified by considerations (as will become apparent: norms and standards) relating to functionality, safety, stability, and the like.

These features, in turn, enable me to study engineering behavior—just like one may analyse the scientist’s behavioral repertoire by relying on the epistemic criteria of table 1.1. More specifically, I will take these artifactual standards as indicative of what engineers ought to do, that is, of norms governing engineering

action.

Finally—and here the parallel arguably ends—we will see that artifactual stan-dards also induce requirements on the part of the artifact’s user. In other words, they point indirectly to what we may call norms of usage.

1.3 Expected outputs

My primary aim is to contribute to the philosophy of technology. I hope to explain, at least partly, what technology is, by analyzing the rules it is played by. I will study the standards the end-products of technology (viz. artifacts) are set against, and how these standards can be (or should be) met. More in particular, my dissertation has two objectives.

First, I want to offer a taxonomy of the different types of norms in the con-text of artifacts. And if we aim for a taxonomy, rather than some kind of arbitrary grab bag, we need some conceptual guidelines allowing us to proceed systemat-ically. More specifically, the conceptual framework should help us to produce a taxonomy that is comprehensive and that contains (more or less) mutually exclusive

categories. As will become apparent, the framework will center around the

distinc-tion between artifact interpretadistinc-tion and artifact creadistinc-tion, and around the nodistinc-tion of successful action, seasoned with a bouquet of epistemic concerns.

My second aim is in fact conceptually prior to the first; indeed, before I can even start my categorization exercise, I need to clarify the chief notion of my project, viz. the notion of “norm”. Now it would be highly convenient if I could borrow someone else’s definition, but unfortunately, though many have consid-ered the issue, philosophers tend to disagree substantially on the meaning of the term. As such, I need to do a considerable amount of conceptual groundwork myself. This groundwork, important to note, is instrumental to the development of the taxonomy; but, as we will see, it might have wider significance. Perhaps we

(26)

can, using artifacts as a test case, point out and repair the deficiencies of existing accounts. Ideally, then, I hope to offer an account of normativity that is useful not only to the philosophy of technology, but to philosophy in general. It is in this sense that the second aim of my project can be considered as an aim in its own right.

1.4 A small discourse on the method

My research aims do not bear witness to the insights of an isolated genius. To understand this, it is convenient to consider briefly the background against which this thesis is set.

When I first mentioned the term “philosophy of technology”, even the fairly well-informed reader might have immediately made associations to renown—or even notorious—philosophers as Marx, Jaspers, Heidegger and Ellul. And indeed, in a certain branch of the philosophy of technology, these authors remain center stage. Recently, for example, when I visited a conference in Galway, out of twelve speakers six reinterpreted the work of Heidegger, two others analyzed Marx’s writ-ings. Although the conference call for papers encouraged analytic thinkers to par-ticipate, a more continental approach is still default. And this approach has largely been, in the provocative words of Joe Pitt, a kind of “social criticism [...], critical de-nunciations of the negative effects of technology on human values and the human life (Pitt, 1999, p.vii).” Put differently: Pitt reproaches traditional philosophers of technology their exclusive focus on the—predominantly negative—impact tech-nology (allegedly) has on individuals and society2_{. Instead, the author argues, we}

should better first center on epistemological and ontological issues. To appraise its impact, it cannot suffice to treat technology as a black box (Kroes, 2000, p. 22); we should open it and engage in a more empirically informed study of actual techno-logical practice, just like the philosophy of science investigates scientific practice, and not only its societal purport.

Without committing myself to any general criticism of the traditional philoso-phy of technology, my project takes Pitt’s and Kroes’s recommendations at heart. In other words, I am interested in “what really goes on in engineering and tech-nology” (Mitcham, 1994, p. ix), and I will therefore try to be as neutral as possible with respect to the question whether the things that actually go on in technology

(27)

1.5. Introduction to the argument and reading guide

happen for the better or the worse. This implies that I, where possible, will draw on empirical material (case studies, evidence from cognitive psychology, design manuals, etc.) to stuff my arguments, and that I try to minimize the use of puz-zling thought experiments and intuition pumps.

One remark, before I give an overview of things to come. As said, artifacts will take center stage in the following chapters, but what kinds of artifacts do I have in mind? I will restrain myself to technical artifacts, that is, material objects that are used and produced to realize practical ends. What I have to say, thus, might but does not need to work for other types of artifacts, such as social artifacts (institu-tions, money, burial practices, etc.), aesthetic artifacts (music, drawings, and other works of art), or unintended by-products of human activity (footprints, debris), nor for undesirable distortions produced by computers (e.g. unexplainable patterns in digital imagery) or by measurement instruments (e.g. unexplainable values in scientific experiments). Excluding these objects certainly does not prevent border-line cases to occur; but concocting a definition with precise boundaries would be enough for an entire dissertation. Moreover, it would presumably require me to defeat a whole army of perhaps necessary, but awkward intuition pumps (does a wooden stick, found in a forest and employed to gauge the depth of a river, count as an artifact? Also when it is used by a chimpanzee? Or by a Martian? Or by a Martian Zombie living on Twin Earth?) I will leave it to others to find a way out of this conceptual morass3. In my analysis, I will therefore only refer to artifacts that I take to be prototypical. And I am confident these have sufficient intuitive appeal to convince even the skeptic.

1.5 Introduction to the argument and reading guide

Unmistakably, the domain I am about to chart is large. To make the assignment more manageable, I will therefore cut it in two. First I approach artifactual nor-mativity from the perspective of artifact usage. In a second step, I analyze it from an engineering point of view.

The underlying idea here might be vague, but has nonetheless sufficient intu-itive appeal: users and designers look at artifacts in a different way. Users want artifacts to serve their practical ends; for designers the practical end is the

pro-3_{Plenty of attempts have been made—with varying success, though. See for instance Hilpinen}

(28)

duction of the artifact itself. Given these different—but clearly related—goals, we can expect the norms to evaluate them to differ as well. Moreover, although the execution of a set of actions is essential to both, using an artifact is prima facie qualitatively different from producing one. Consequently, it would not come as a surprise if we were to discover that norms governing usage and design are not alike.

Whether this distinction cuts reality at its joints is not my primary concern; the bifurcation is rather a heuristic to get the argument going. Its merits—if any— thus should surface only in the course of the argument, or even later, in the end-product, viz. the taxonomy.

The basic structure of the thesis, now, reflects the dual perspective of use and design. In particular, Chapters 2 to 5 cover norms related to artifact use; Chapters 6 and 7 take up the normative features of artifact design. Finally, in Chapter 8, I forge a taxonomy using the elements gathered along the way. In what follows, I give more substance to the arguments of the individual chapters.

Chapter 2 introduces us to the realm of artifact usage. In particular, it inves-tigates a basic phenomenon called artifact hermeneutics, that is, the way people attribute functions to artifacts. To get the argument off the ground, I rely on Daniel Dennett’s writings. I show the deficiencies of his account, and formulate an alter-native. At the end of the chapter, we should have a clear grasp of the cognitive why’s and how’s of artifact interpretation.

The account offered in Chapter 2 is descriptive. It explains how people ascribe functions to artifacts, but is neutral with regard to the correctness of these ascrip-tions. That is where Chapter 3 comes in. I explain that appropriate attributions should be backed up by a use plan and context of use that enable the successful realization of the end the artifact is said to serve. This suggestion, however, does not do right to the notion of proper function yet. Indeed, a screwdriver can be used as a can opener, implying there is for the object a use plan and context of use promoting the goal of opening cans; but although it might be appropriate to inter-pret the screwdriver accordingly, such an interinter-pretation is hardly default. Proper function and use, I argue, do not only require (counterfactual) successful action, but also some kind of institutional support.

Subsequently, the notion of successful action can be deployed to categorize artifactual norms related to artifact usage. But first, Chapter 4 needs to settle one last terminological issue. In particular it furnishes a characterization of norms and

(29)

1.5. Introduction to the argument and reading guide

normativity. Against standard approaches, I claim that a norm is a comparative measure not only of how things ought to be done, but also of how things (s.a. artifacts) ought to be.

Groundwork done, Chapter 5 offers an overview of the norms induced in arti-fact interpretation and use. As said, the notion of successful action will guide us here. More precisely, I will show that successful action poses requirements on the artifact’s user, and on the artifact itself. Furthermore, I assess the normativity of these requirements, relying on the conceptual framework developed in Chapter 4. With Chapter 6 we cut into the second part of the thesis and switch to the perspective of artifact design. Using a slightly different conception of success-ful action, I explain that some artifactual norms are not—or only marginally—to the concern of users, but which designers (should) have a particular interest in. These norms relate to the different phases in an artifact’s lifecycle; quality entails meeting the requirements set in each of these phases.

Subsequently, Chapter 7 analyses how these norms should be attended to by artifact designers. As such, the chapter concerns norms on actions, in particular, norms for doing design. From a case study I distill two broad categories: codes of good practice and ethical codes.

Finally, Chapter 8 compiles the distinct normative categories considered through-out the thesis. I give some final comments on the resulting taxonomy and rehearse the outlines of my overall argument. Furthermore, I explain the significance of my work for the philosophy of technology.

Perhaps some reading advice comes in handy. For novices to the philosophy of technology: Chapter 2 is the most entertaining chapter, an appetizer that well might provoke further interest in the subject. For professional designers: the tax-onomy of Chapter 8 represents the norms I believe to govern your profession. For philosophers in general: the most interesting arguments, philosophically speak-ing, can be found in Chapter 3 and 4. For philosophers of technology: since you are my target audience, I hope you to read the entire manuscript.

(30)

(31)

Chapter 2 The interpretation of artifacts

1

2.1 Introduction

Human beings are proficient in operating pens, pocket calculators, paper, paper-clips, television sets, cars, microwave ovens, the bottles they drink beer out, auto-matic adjustable waterbeds and so forth. They understand that clothes are to keep them warm, that soap is useful to wash their hands, they know how to turn lights on and off and successfully figure out their word processor’s automatic numbering habits so as to know how to change them. Failing to understand the purposive-ness of man-made objects comes close to failing to live in an environment that is pervasively artifactual.

Indeed, language and tool use are routinely cited as the paradigmatic behaviors involved in high-level cognition; both are sometimes even considered as typically human, as qualities distinguishing us from other animals (Preston, 1998). But whereas language has received a great deal of attention in the philosophical lit-erature, philosophers surprisingly have shown little interest in our ability to use tools2_{. Exceptions can be found in recent discussions on functions}3_{, which mostly}

concern the functional aspects of biological and, indeed, artifactual objects (see for instance Wright, 1973; Cummins, 1975; Millikan, 1984, 1989; Neander, 1991a,b;

1_{Much of the material presented in this chapter can also be found in (Vaesen and van Amerongen,}

2008).

2_{Dipert (1993) also notes this lacuna, and suggests some possible reasons for it.}

3_{The discussion was initiated in the philosophy of science, in particular by Hempel (1959) and}

Nagel (1961). They demonstrated the significance and logic of function talk for the social sciences, which was later applied to other sciences, such as biology.

(32)

Griffiths, 1993; Godfrey-Smith, 1993, 1994; Kitcher, 1993; McLaughlin, 2001; Ver-maas and Houkes, 2003). But as Lewens (2004) notes, even when these authors consider artifacts, it is in an attempt to define the meaning of the concept “func-tion”. Take for instance what Larry Wright (1973, p. 161) writes in his seminal paper “Functions”(concerning both biological items and artifacts):

The function of X is Z means (a) X is there because it does Z,

(b) Z is a consequence (or result) of X’s being there.

In fact, as Godfrey-Smith (1993) points out, much of the literature since 1973 has engaged in the refinement of Wright’s original ideas. Most writers, like Wright, involve in a conceptual analysis of “function”, in an attempt to define the sufficient and necessary conditions for the correct application of the concept.

For the sake of bringing clarity in the inconsistencies and the impreciseness of natural language concepts, conceptual analysis might be indeed a useful tool. But it is too optimistic to suppose that once we have established “the” meaning of the concept of function, we know everything that there is to know about the human ability to understand and operate artifacts. Ideally, then, we would appreciate what people mean—or ought to mean—when attributing a function to an artifact, but we would remain ignorant on, so to speak, the cognitive why’s and how’s of such an ascription. I take it that we can study the human animal’s artifactual behavior without focusing exclusively on its linguistic representations; after all, that is how one would study the behavior of other animals, such as primates which are found to display primitive forms of tool use.

Therefore this chapter will develop a descriptive account of how people form beliefs about artifacts. I will argue that they ascribe functions relying on three considerations, namely optimality, original intent, and usage. In other words, sometimes an artifact’s function is understood as its optimal function, sometimes as its intended function, sometimes as the function the artifact is used for.

Since I will develop a descriptive account, for the moment I will not worry about the appropriateness of people’s functional beliefs; I merely want to find out how they are formed. In the next chapters, then, we turn to some normative and epistemic issues. Furthermore, although the notion of “function” is crucial to our understanding of artifacts, other notions too play a central role, and I will consider

(33)

2.2. Dennett’s optimality considerations

these in due course. But for the moment, I take, as most artifact theories have done thus far, functions center stage.

2.2 Dennett’s optimality considerations

Although philosophers paid little attention to our capacity to grasp the purposive-ness of man-made, designed objects, there is one exception that will serve as a starting point for my analysis, namely the work of Daniel Dennett. In several writ-ings Daniel Dennett (see Dennett, 1978, 1989a, 1990, 1995) has considered arti-fact hermeneutics, that is, the way people discover what an artiarti-fact can or should be used for. To explain how they do it, Dennett relies on the notion of the design stance, a stance “where one ignores the actual (possibly messy) details of the phys-ical constitution of an object (Dennett, 1989a, p.16-17)” and determines what it is for under the assumption that it is optimal. The latter means one assumes both that the explanandum has a capacity to fulfill the role it has been designed to—i.e. it does not malfunction—and that its designer did not do anything in vain, that is, every component of the object has a raison d ’être and contributes to the object’s well-functioning (see for instance Dennett, 1995, p. 212-213). The reasonableness of a function ascription should be set against these constraints; to determine the function of an artifact we look at what the artifact would best be able to do. If we find an artifact that would be perfectly able to, say, pit cherries (and nothing else), it is a cherry-pitter. Or as Dennett writes:

It counts against the hypothesis that something is a cherry-pitter, for instance, if it would have been a demonstrably inferior cherry-pitter. (Dennett, 1990, p. 184)

It is a cherry-pitter, no matter what the object’s designer or its users intended it to be. In other words, function interpretation is not an interpretation of intentions:

So what something is really for now is no more authoritatively fixed by the current user’s “intentions” than by any other intentions. (Dennett, 1990, p. 194)

Dennett’s suggestion displays striking similarities to an idea that was already dominant in the literary criticism of mid twentieth century—the so-called New

(34)

Criticism. In their seminal paper “The Intentional Fallacy” Wimsatt & Beardsley for instance argue that:

the design or intention of the author is neither available nor desirable as a standard for judging the success of a work of literary art. (Wimsatt and Beardsley, 1954, p. 3).

Instead the meaning of a poem should be assessed in terms of optimality: Judging a poem is like judging a pudding or a machine. One demands that it work. It is only because an artifact works that we infer the intention of an artificer. (Wimsatt and Beardsley, 1954, p. 4)

It is obvious what kind of literary criticism Wimsatt cum suis is attacking: tra-ditional theories that take as primary focus the meaning intended by the author. Regarding the interpretation of artifacts, Dennett attacks a similar audience; he opposes those who refer to designer’s intent to answer the question what an ob-ject is for. As such he fights a seemingly common-sensical intuition, one that has inspired the accounts of Dipert (1993), Vermaas and Houkes (2003), and most clearly McLaughlin (2001), who writes:

An entity is an artifact and has a particular artifactual function if it is assembled, reassembled, or virtually reassembled with that particular purpose in mind. (McLaughlin, 2001, p. 55)

As van Amerongen (2008) notes, Dennett has several reasons to adopt an anti-intentionalistic strategy. First, it fits well with his insistence on the indeterminacy of intentional interpretations. Inscrutability of interpretation regarding human behavior means that there are no deeper facts that can settle the matter unequiv-ocally. And to be clear, it regards any kind of human behavior, including artifact creation. Not even the designer herself may know what exactly her intentions were when creating an artifact:

The inventor is just another user, only circumstantially and defeasibly privileged in his knowledge of the functions and uses of his device. (Dennett, 1990, p. 186, italics added)

So, intentions are not reliable indicators of function, for the simple reason that they are unreliable indicators, full stop.

(35)

Second, designer intentions do not fix an artifact’s function, since no artifact is immune to losing its original function. Indeed, old-fashioned sad-irons may be turned into bookends and many European churches are nowadays used exclu-sively to host non-religious events. In such cases, we cannot infer from intended function to current function. Put differently, designer’s intentions are simply ir-relevant.

Finally, Dennett defends a notion of function devoid of intentionality, because he wants it to apply to both technical and biological items. Even stronger, he aims to establish that the interpretation of people, biological traits, texts and artifacts are entirely “the same project addressed to different objects (Dennett, 1990, p. 177).” For a severe anti-creationist like Dennett, biological functions are not intentionally designed. Hence, for his generic account to work, intentions cannot be taken central .

These three reasons might for Dennett be sufficient to reject intentional ap-proaches, but the author nowhere offers a real positive argument in favor of his alternative, viz. a design stance relying on mere optimality considerations. In-stead, its plausibility is illustrated by a number of examples. To see the design stance at work, Dennett refers for instance to reverse engineering:

When Raytheon wants to make an electronic widget to compete with General Electric’s widget, they buy several of GE’s widgets and pro-ceed to analyze them: that’s reverse engineering. They run them, bench-mark them, X-ray them, take them apart, and subject every part of them to interpretive analysis: Why did GE make these wires so heavy? What are these extra ROM registers for? [...] Notice that the reigning assumption is that all these "why" questions have an-swers. Everything has a raison d ’être; GE did nothing in vain.[...] If the reverse engineer can’t assume that there is a good rationale for the features they observe, they can’t even begin their analysis. (Dennett, 1995, p. 212-213)

Moreover, Dennett mentions archaeologists and historians interpreting the ar-chaeological record. These have for instance discovered the Antikythera mecha-nism to be an orrery or a planetarium, and the “proof of that is that it would be a

good orrery (Dennett, 1990, p. 184).” One has calculated the periods of the rotation

(36)

repre-sentation of what was then known about planets and their orbits. If so, Dennett argues, the function attribution—which is almost certainly correct—came about without relying on original intent. Rather, it was established via an examination of the purposes the unknown object optimally could have served.

Both arguments are descriptive: Dennett pictures the real-life practice of en-gineers, archaeologists and historians, that is, the way they involve in functional reasoning. Elsewhere however, the design stance seems more like a methodolog-ical advice. When interpreting a robot, for instance, Dennett recommends us to consider it “in the context of the costs and benefits of its current environment”, since in this way “we may arrive at a better interpretation of its internal states than its original designers can muster (Dennett, 1990, p. 186, italics added).” Else-where we read:

Curiously, then, we get better grounds for making reliable functional attributions [...] when we ignore “what people say” and read what func-tion we can off the discernible prowesses of the objects in quesfunc-tion, rather than off the history of design development. (Dennett, 1989a, p. 319)

For the rest of the chapter I will set the normative reading aside, though, and focus exclusively on the descriptive merits of Dennett’s theory4. In short, the latter espouses the idea that in artifact hermeneutics people reason in terms of optimality, rather than in terms of designer’s intentions. So the function of a certain item is its optimal function, not the one its designer intended it to be.

Before moving on, though, I need to clarify one last ambiguity in Dennett’s proposal. Artifact hermeneutics, as said, is about discovering the function(s) of an artifact. But what exactly does this mean—at least according to Dennett?

Given the importance Dennett attributes to optimality, we can dismiss two in-terpretations from the outset. According to the first, artifact hermeneutics aims to settle the question: "What does artifact X afford?" In other words, it is the artifact’s possible function(s) we are after. But possible functions are, per defi-nition, all those that an artifact is able to perform independent from some sort of

optimality standard; we only require the artifact to have the capacity to realize the

function(s) in question. For instance, Dennett is not interested in all the possible

4_{For more information on the design stance as a normative theory see Vaesen and van Amerongen}

(37)

function ascriptions archaeologists may produce with respect to the Antikythera mechanism—after all, the object may be used for a wide variety of things, e.g. as a hammer, a paperweight, an electric conductor. Rather, the author aims to ex-plain how archaeologists derive one (i.e. the most salient) interpretation which establishes "the" (alleged) function of the device.

According to the second implausible interpretation, artifact hermeneutics ad-dresses the question: "What does artifact X afford right now?" In this case, we are interested in the device’s current function. But again, optimality considerations are redundant in this regard. Determining what artifact X is currently used for can be done without us needing to consider whether the device is used for the things it is best able to do. Seeing an archaeologist using the Antikythera mech-anism as a paperweight provides me with perceptual evidence for the artifact’s current function. I may appreciate that such usage is suboptimal, but my finding does not affect (the reasonableness of) my original interpretation.

In a more plausible reading, artifact hermeneutics concerns—for Dennett at least—the question: "What function is artifact X meant to afford?" If so, Den-nett addresses, somewhat paradoxically, the issue of discovering intended func-tion. But—and this is the controversial element in Dennett’s proposal—to infer intended function, we should not take intentions as a starting point, but rather optimality. For instance, the archaeologist looks for the purposes the Antikythera mechanism might have served for the people once producing and/or using the object, in short, for the device’s alleged intended function; and to determine it, she will sort out what the Antikythera mechanism is best able to do, taking this ca-pacity indicative of what the device was meant to afford. Or more generally, to sort out an artifact’s intended function, we should look at what the object in question affords best, and not at what people say or have said about it.

Note moreover that intended function is not always equal to original intended function; what an artifact is meant to afford may be determined by its designer(s), but also by its users. In the latter case, the intended function of say, a sad-iron being turned into a bookend, is altered, whereas its originally intended function remains intact. Given Dennett’s treatment of the sad-iron example, we may infer that the author discusses the meaning of artifacts irrespective of who (i.e. designer or user) it was that meant them to afford this or that. In other words, the author is concerned with intended function in general, not merely originally intended function.

(38)

But, one might argue, doesn the sad-iron example not show that, in contrast to what we claimed above, Dennett’s main concern is the interpretation of possible, or perhaps current function? If we interpret the sad-iron as a bookend—whether or not relying on an optimality assumption—it indeed seems we are proposing one of the iron’s possible (and current) functions. There is a subtlety in the exam-ple, though, that weakens this line of reasoning. Dennett is not talking about one individual idiosyncratically using the iron as a bookend, rather about a group of people appropriating, by convention so to speak, the object for that purpose. So there is still some kind of privilege given to the bookend ascription. It has been promoted from a mere possible function to the status of (apparent) intended func-tion; through their repeated use as bookends, sad-irons are nowadays meant to afford upholding books—or so might Dennett argue.

In sum, I believe that it is more plausible to assume that Dennettian artifact hermeneutics concerns intended function, rather than possible or current func-tion. In the following three sections (sections 2.3, 2.4, 2.5), then, I will show that to settle the question "What is artifact X meant to afford?", people—contra Dennett—do not merely rely on optimality standards, but also take intentional considerations aboard. The results will help us to force an entry into the realm of norms governing artifact interpretation and use.

2.3 Artifact hermeneutics and cognitive psychology

If the design stance is to count as an empirical claim, it should give us an accurate description of how real-life people commonly reason about artifacts. But in fact, it does not. To see why, it is instructive to have a look at recent research in cog-nitive psychology, where one has studied how children infer what artifacts are for. Roughly, the research can be subdivided in two research traditions, an essential-ist and an anti-essentialessential-ist one. And although these disagree on some points, the results of both can be used to challenge Dennett’s account.

Let us start with the essentialist tradition. Many scholars working in this tra-dition have, directly borrowing from Dennett, labeled the human strategy to in-terpret man-made objects the “design stance” (see for instance Kelemen, 1999a,b; Defeyter and German, 2003; German and Johnson, 2002; Kelemen and Carey, 2006). Surprisingly, though, these authors have given the term quite a different meaning: for them the design stance refers to the cognitive capacity to

(39)

catego-2.3. AH and cognitive psychology

rize and reason about artifacts on the basis of the original intentions with which these objects were produced (and thus not in terms of Dennett’s non-intentional optimality considerations). In fact, this idea is the core idea of the essentialist program: people reason about artifacts according to an “essence”, namely origi-nal intent. This means that origiorigi-nal intended function overrides such factors as appearance and context and actual use as the basis for how humans categorize artifacts.

It is around the age of five or six that the design stance—in the sense defined above—gets mature—at least according to authors such as Kelemen (1999b); Ger-man and Johnson (2002); Defeyter and GerGer-man (2003); Matan and Carey (2001). One way to test their intentional bias proceeds by confronting subjects with objects having a clear original intended function, but used in an improper way. Kelemen (1999b), for instance, presented people of different age with a picture of an arti-fact. The subjects were told that the object was made by a person who intended it to be a thing to stretch clothes—the person needed such a tool, since his wardrobe got shrunk by his washer. Moreover, they were informed that the artificer, after a onetime use, gave the artifact to a friend, and that the friend used the thing to stretch and exercise her bad back. When asked what the object was for, subjects predominantly answered it to be for drying clothes, in other words, its function was equated with the one intended by the artificer. Hence, assuming the artifact to be optimal for both tasks, there seems to be a bias towards original intent when people categorize artifacts—pace Dennett. For if people would not take design in-tentions into account, we would expect an equal distribution of fifty to fifty in the responses given.

The results of another study, conducted by Gelman and Bloom (2000), point in the same direction. These scholars showed that whether the very same crude object was described as the product of intentional design or as the outcome of an accident affected how it was interpreted. For instance, a paper hat was more frequently interpreted as a hat, if subjects were told that it was intentionally de-signed for that purpose, than when they were informed it was the result of an accident—c.q. the result of a car running over a newspaper. If optimality would be as essential as Dennett thinks, we again would expect the object, having the same “optimality ” in both scenarios, would be interpreted as often as a hat in both the intentional and accidental case; the apparent discrepancy in numbers, though, is explained by the fact that the subjects took intentional considerations

(40)

aboard.

Another particularly interesting study, still essentialist in nature, was con-ducted by Kemler Nelson and colleagues (see Kemler Nelson et al., 2002, 2004). The team investigated how people interpret broken objects. To see in what sense their efforts challenge Dennett, one should first recall what his optimality assump-tion amounts to. First of all, it means that one has to assume that the object one interprets is not malfunctioning, and second, that all of its components have a rai-son d’être. It is the first condition which interests us here; non-malfunction must be the default assumption, otherwise it is impossible to get any interpretation off the ground. With this in mind, let us turn to Kemler Nelson and colleagues. They showed that adults (and children, from 10 years on) interpret accidentally mal-functioning objects according to original intended function. Subjects, for instance, were presented with a damaged cup (the thing could not hold liquid because an ir-regular piece of its side was broken off); most of the participants nevertheless still interpreted the object as a cup. Similar results were obtained for interpretations of less familiar artifacts, such as "beckets" (a ball dispenser), "luzaks" (a tool to draw circles), and other objects specially designed for the experiment. Now, even if people would typically reason in terms of optimality, like Dennett argues, the results of this study show that humans at least at a certain point give up optimal-ity; instead of continuing to look for a function the broken thing still can perform optimally, they attend to the intentions with which it was designed. And as impor-tant, doing away with the non-malfunction condition did not prevent reasonable interpretations getting off the ground.

Before we consider the anti-essentialist research tradition, one remark. On some occasions, as explained before, Dennett seems to defend the claim that what an artifact is for may change over time, and may for instance be determined by its users—sad-irons interpreted as bookends being a case in point. Although there undoubtedly is a flexibility (e.g. over time) in the interpretations people make, we should be wary to overestimate its significance. Humans appear fairly conserva-tive; cognitive psychologists, for example, have argued that children, already at the age of six or seven, start to show the signs of “functional fixedness ” (see e.g. De-feyter and German, 2003; Casler and Kelemen, 2005). They tend to become “fixed ”on the original intended function of an artifact; and it is particularly difficult for them to reason otherwise. They not only “theoretically ” frame an artifact in terms of original intended design, but will actively and practically use this knowledge

(41)

2.3. AH and cognitive psychology

(or bias, if you wish) when asked to solve specific problems with the artifact. For instance, a study of Defeyter and German (2003) showed that six-seven year old children were slower than five years old in solving a problem by using an artifact in an atypical way.

There are no clear-cut answers as to why people develop such a bias. Casler and Kelemen (2005) hypothesize that thinking in terms of original intended function provides us with a helpful shortcut in our daily interaction with artifacts; it would be highly impractical if we needed to ponder over and over again on the function of all the man-made objects we encounter. As such, it is suggested that reasoning in terms of original intent may have a place in our evolved human nature—a fact that, if correct, Dennett as a naturalist should be glad to accept.

Scholars in the anti-essentialist tradition are more pragmatic; they believe that humans categorize artifacts not according to some essence (such as original in-tent), but rather depending on the task at hand.

Some researchers for instance have discovered a phenomenon known as the shape bias. They argue that, when people are asked to judge an object’s kind, phys-ical features are sometimes given more weight than original intent (see e.g.Malt and Johnson (1992); Landau et al. (1998); Hampton (1995); Baldwin (1992); Gra-ham et al. (1990); Sloman and Malt (2003); but see Gelman and Bloom (2000); Diesendruck et al. (2003)). Chaigneau (2002), for example, considered cases where people have privileged access to the artificer’s intentions by being told. He pitted such knowledge against other aspects of the object and considered their rel-ative importance in judgments of the appropriateness of a label. For instance, in one scenario the experimenter intended a certain object to be a mop and used it as a mop, but the object was a bundle of plastic bags attached to a four foot long stick. Most subjects did not regard the object a mop, despite the artificer’s intentions and despite the fact that it was used as such. Chaigneau concludes that physical affor-dances and context of use may well dominate original intentions when subjects reason about artifact categories and artifact functions.

In a similar vein, Siegel and Callanan (2007) show that when subjects are told that many people (as opposed to one individual) use an artifact in a new way, they not strictly interpret the artifact in terms of its intended purposes; put differently, their interpretations are sensitive to apparent conventions of use. As such, the researchers criticize the “essentialist ” design stance for neglecting the ways in which tools are used in social settings.

(42)

Note, however, that both Chaigneau and Siegel & Callanan do not claim that original design intentions are simply irrelevant to function ascription. On the con-trary, the authors recognize that it is an important feature, but add that it is not the only one. For Chaigneau, physical affordances and context of use may well dominate original intentions, depending on the case at hand. Likewise, Siegel & Callanan recognize that conventional functions of artifacts are typically related to their original intended function, and that this may account for the stability in our reasoning about artifacts—in other words, for our functional fixedness. Never-theless, the authors present convincing evidence for the fact that, given its embed-dedness in a social context, an artifact’s meaning may change and evolve over time and culture.

So in sum, whether one adheres to an essentialist or to a more pragmatic view, it is safe to say that, according to cognitive psychology, designer’s intentions play an important role in the interpretation of artifacts—the main topic of disagree-ment being exactly how important. And obviously, this poses a problem for the design stance as Dennett pictures it. As an empirical theory it is in conflict with a converging body of scientific evidence; it misrepresents how people usually frame artifacts. In particular, contra Dennett, we have explained that to decide what an artifact is for, people (often) reason in terms of intentions—or, to use Dennett’s words, in terms of what “people say ”—not (only) in terms of what it the object is best able to do.

2.4 Artifact hermeneutics in archaeology

Dennett then might perhaps claim that the design stance is not so much an em-pirical theory about our everyday functional reasoning, but is meant to capture a methodology used in the scientific interpretation of artifacts. This strategy will not work, however.

What is the proper domain to evaluate scientific artifact hermeneutics? Many scientific fields use a notion of function, especially in the humanities and the social sciences. But studies of artifact function are rare. The only exception is archaeology—indeed, a domain frequently considered by Dennett himself. The archaeologist tries to find out the function of an artifact, and tries to do so in a scientifically respectable way.

(43)

2.4. AH in archaeology

the archaeological record: what the “designer” of a prehistoric artifact intended is not given, and worse, it is often irrelevant in the sense that we are not even sure whether early tool-makers had intentions at all. Another factor in support of Dennett is that the archaeological literature frequently refers to optimality consid-erations. Take for instance Salmon’s rule of thumb for the archaeologist:

The more severe the limitation on the form of an object that the sus-pected function imposes, the more reliable is the ascription of that function [...] For some objects, such as grinding stones, there are very few forms that are compatible with reasonably efficient performance of the function. (Salmon, 1982, p.59)

The principle is relevantly similar to Dennett’s optimality principle that it would count against something being a cherry-pitter, if it would be a demonstrably infe-rior cherry-pitter. Nevertheless, there are a number of cases and questions that seem only resolvable by appeal to intentionality. Let us consider an example.

Near Baghdad many so-called Babylonian batteries have been found, objects belonging to the Parthian (between 250 BC and 224 AD) or Sassanian era (224-640 AD). Each so-called battery is a 15cm vessel which contains a cylinder of sheet copper, capped at the bottom, in turn covering and protecting an iron rod, which shows signs of acid corrosion (see figure 2.1).

In 1938 Wilhelm König, director of the National Museum of Iraq, published a paper suggesting that the artifact may have been a galvanic cell, a kind of primitive battery (König, 1938; Dubpernell, 1978). And that would indeed be a spectacular discovery if it were true. It would imply that electrical current had been used by the ancients and was only rediscovered by Galvani and Volta, some 1,800 years later.

König’s hypothesis was tested by (Jansen et al., 1987). They concluded that, when fueled with a solution of benzoquinone, a substance occurring naturally in the secretions of some beetles, the object could indeed produce a certain volt-age (+/-0.87V); in other words, physically the artifact can fulfill the function of a battery. The question—to Dennett—then is: is this sufficient to call the design optimal, and thus to conclude the thing to be a galvanic cell?

A “yes” would be—and has been—whole-heartedly embraced by proponents of the paranormal. Some of them (see for instance de Montellano, 1991; von Däniken, 1968) find in the “power source” thesis evidence for a technologically

(44)

ad-Figure 2.1: A so-called Babylonian Battery

vanced (extraterrestrial) civilization in remote antiquity. Most historians, however, are fairly sceptical (see Eggert, 1996), since the alleged galvanic cells are contem-porary with the growth and height of the Roman Empire. The latter is a fairly well-documented era, hardly a period thus in which such a civilization would have gone unrecorded, particularly since the Parthian Empire was Rome’s principal enemy in the east. Furthermore, what kind of appliances would need the energy supplied by these cells? Until now, no Parthian electronic devices have been found, so at the face of it, there was no Parthian desire—and hence no intention—to produce electricity whatsoever.

The hypothesis that the Babylonian vessel was used as a storage device for sa-cred scrolls, which were wrapped around the iron rod—as, among others, Paszthory (1989) suggests—seems more plausible—and more accepted—than the “power source” hypothesis, but not unequivocally in the light of optimality. Indeed, the vessels are good at storing parchment or papyrus, but why would they need an iron rod and a copper cylinder (and an asphalt seat at the bottom and an asphalt stopper on top)? In engineering terms, the artifact is over-designed; and thus it

(45)

2.4. AH in archaeology

would be in conflict with Dennett’s optimality principle which states that every ar-tifact’s component should have a raison d’ être. Besides, if one thinks over-design is unproblematic, why not just claim the artifact to be a container full stop, or a container of air? Because without the restriction of over-design, such a hypothesis would be just as reasonable as the “power source” hypothesis.

More important for our purposes, to come to his conclusions, Paszthory rea-sons in terms of intent, not mere optimality. He explains iron and copper to have a magical meaning, playing a role in ancient alchemy. According to the Parthi-ans, it might well have been that the use of these metals was meant to please the gods, an attempt to protect their sacred documents against divine terror. As such, Paszthory hypothesizes about what Parthians could and could not have meant the artifact to be for—inferring from knowledge about their beliefs and desires—to adjust his interpretation accordingly. The example illustrates that, pace Dennett, a methodologically sound interpretation of an artifact may well involve an interpre-tation of intentions, beliefs and desires.

Successful experiments in experimental archaeology, as Eggert (1996) remarks, can only show a supposed ancient technique to be possible, but never its applica-tion. For instance, when Thor Heyerdahl crossed the Atlantic in an Egyptian boat, he only showed that it was possible, in principle, for the Egyptians to have done likewise. But to accept the claim that they indeed did, one would need archaeo-logical evidence from America. Similarly, one should be wary of interpreting the experimental evidence about the Babylonian vessels as decisive. Even if the ves-sels can generate current—perhaps even optimally—this does not prove that they once did. In contrast, the example shows to what kinds of anomalies optimality conditions can lead, if not constrained by some kind of intentional reading; I sin-cerely do not believe extraterrestrial visitors once learnt the Parthians to produce energy. And if they did, they were remarkably parsimonious; being visitors with space travel capabilities, why did they not show the Parthians less primitive ways of producing energy?

In fact, these remarks also apply to the examples Dennett himself gives. For instance, Dennett considers William Calvin’s analysis of the so-called Acheulian hand axe (Calvin, 1986). The object is an artifact mainly found in Ethiopia. It is a flattish stone, arising about 1.8 million years ago (see figure 2.2).

For a long time it was thought that the object was a kind of hand axe, a thing used for instance for butchering. But actually, it would be a fairly poor hand axe.

(46)

A philosophical essay on artifacts and norms

A philosophical essay on artifacts and norms

A Philosophical Essay

on Artifacts and Norms

Krist Vaesen

‘Wonder en is

gheen wonder’

Contents

Acknowledgements

Chapter 1

Introduction

1.1

A Chinese experiment

1.2

An instructive parallel

1.3

Expected outputs

1.4

A small discourse on the method

1.5

Introduction to the argument and reading guide

Chapter 2

The interpretation of artifacts

1

2.1

Introduction

2.2

Dennett’s optimality considerations

2.3

Artifact hermeneutics and cognitive psychology

2.4

Artifact hermeneutics in archaeology